*Cluster configuration

Owned by Tom Polman
Last updated: Feb 27, 2024 by Martijn Friggen
4 min read

Configuration Article | CA-20220531-MS-01

VDG Sense | OpenAPI |

Introduction

The cluster setup allows all joined servers to distribute the configuration data over all servers and required to create a multi-server installation.

VDG Sense 2.5 – Management/Slave setup

Management/Slave

  • Macros are executed cross servers

  • User management files stored on the management server

  • Each server maintains its own device configuration

Failover

  • Hot standby server for failing hardware

  • Maintains a copy of each server configuration

VDG Sense 2.6 – Cluster setup

Cluster

  • Distributed user and roles configuratios

  • Each physical server has all users and roles

  • Management server holds serverlist, layouts and macro configuration

  • Client needs to login on Management server

Failover

  • No need to maintain configuration backups

Upgrading 2.5 to 2.6

When updating a VDG Sense installation from 2.5 to 2.6, it’s important to note that a multi-server installation needs some additional manual configuration before it fully functions again. Each server in the installation need to join the new cluster setup. This also includes the Failover server, which now needs to be part of the server list.

Joining a cluster

The cluster configuration is done on the management server of the cluster. This is the ‘main’ database that will link all secondary servers. To configure the cluster, follow these steps:

 

  1. Login on the Management-server

  2. Click on the configuration wheel to open the settings

3. Navigate to the “Servers” tab

4. Add the secondary/slave server(s) to the server list. This can be done either manually or through discovery

5. Select the secondary server in the server list

6. In the “General” subtab, click the “JOIN” button in the Cluster settings

7. Enter the IP and port (default 443) of the management server in the dialog

8. Click OK

9. Repeat for all secondary servers

Leaving a cluster

Leaving a cluster means that the server will no longer act as a slave server. This means that created user and user roles will be removed. It’s possible to leave a cluster from both the management server, or directly from the slave server, but if leaving the cluster from the slave server, it’s not possible to rejoin a cluster right away.

To leave a cluster, follow the following steps.

  1. Login on the management server (recommended) or the slave server directly.

  2. Click on the configuration wheel to open the settings

3. Navigate to the “Servers” tab

4. Select the secondary server in the server list

5. In the “General” subtab, click the “LEAVE” button in the Cluster settings

7. Confirm leaving the cluster in the dialog

8. Click Yes to confirm leaving the cluster

 

For more information, see Failover setup article

Cluster status

The cluster can have the following status

Cluster Management

The current selected server contains the management cluster or has not joined any cluster and is self-contained. This means that the server is either the management server, or an unconfigured slave server.

Cluster Joined

The current selected server has joined the cluster setup. This status is only visible on slave servers or the Failover server.